Participants
Table 2. Characteristics of the sample (n=22).
|
min
|
max
|
M
|
SD
|
Age
|
65.00
|
80.00
|
74.50
|
3.64
|
Height, m
|
1.54
|
1.85
|
1.71
|
.10
|
Weight, kg
|
60.00
|
103.00
|
76.18
|
12.54
|
BMI, kg/m2
|
20.75
|
32.51
|
26.10
|
3.52
|
HRresting, bpm
|
51.00
|
85.00
|
63.55
|
7.62
|
HRmax, bpm
|
161.87
|
173.69
|
166.01
|
2.54
|
HR40 %, bpm
|
95.35
|
117.47
|
104.53
|
4.86
|
HR60 %, bpm
|
117.52
|
133.70
|
125.02
|
3.64
|
Abbreviations: min minimum, max maximum, M mean, SD standard deviation, BMI Body Mass Index, HR heart rate
|
The results of 22 participants with an age of 74.50 ± 3.64 years were evaluated (Table 2). Nine participants were male and 13 were female. All participants had essential hypertension as per ICD-10-GM-2021 I10. Most of them were taking at least one of the following drugs: angiotensin II receptor blockers (n=9), anticoagulants (n=7), calcium channel blockers (n=5), ACE inhibitors (n=3), diuretics (n=2), angiotensin-converting enzyme inhibitors (n=1) or β-Acetyldigoxin (n=1). Five participants were taking β-Blockers. Three participants were not using any medication for the treatment of hypertension. One person reported occasional dizziness in daily life. Ten of the participants had previous VR experience, gained professionally (virtual built house) through other research projects or in museums and exhibitions. Sixteen participants reported being active in sports more than once a week. Six participants were active in sports less than once a week. On average, the participants participated in sports 2.5 times per week (including strength training, dance classes, Pilates, cycling, rehabilitation sports, water gymnastics, meditation, Nordic walking, jogging, swimming, horseback riding and hiking). None of the participants had cognitive impairments, measured by the TICS (37.27 ± 2.57 points) or an increased risk of falling, measured by the Tinetti test (27.64 ± 0.79 points).
Load intensity
The evaluation of the load intensity was based on the continuously recorded HR during the VR exergame. For this purpose, the mean values of the HR values of the different training sections were compared with each other (Table 3). Significant differences between the two training sessions were found in the mean value of the “main part” as well as at the “end”. In addition, the maximum HR achieved during training differed significantly between the VR-SET and the VR-ET.
Table 3. Mean HR values of the VR-SET and VR-ET.
|
VR-SET
M (SD), bpm
n = 21
|
VR-ET
M (SD), bpm
n = 21
|
p-value
|
95 % CI
|
d
|
Start
|
86.14 (14.12)
|
83.05 (14.27)
|
.117
|
-.84, 7.03
|
.358
|
Warm-up
|
94.01 (16.99)
|
93.04 (14.02)
|
.686
|
-3.93, 5.86
|
.090
|
Training phase
|
106.08 (19.70)
|
96.98 (15.16)
|
.000***
|
4.84, 13.36
|
.973
|
Cool down
|
104.42 (20.25)
|
103.51 (15.64)
|
.687
|
-3.76, 5.57
|
.089
|
End
|
100.67 (19.73)
|
93.33 (15.05)
|
.002**
|
3.14, 11.52
|
.797
|
HRmax
|
124.90 (20.65)
|
114.48 (14.34)
|
.000***
|
5.93, 14.93
|
1.055
|
Paired t-test, M mean, SD standard deviation, CI confidence interval, **p < 0.01, ***p < 0.001.
|
As the mean values of the “main part” showed, higher loads were achieved in the VR-SET than in the VR-ET. In the “main part” of the VR-SET, half of the participants reached their personal intensity range of at least 40% (lower limit of the moderate target zone). In contrast, in the VR-ET, only five participants (27.73%) managed to reach a load intensity of at least 40% (Fig. 6).
Fig. 6. Progression of mean HR during training phase in VR-SET and VR-ET per participant. The black circles indicate exceeding an individual load intensity of 40 %. Participant 8 in Fig. B is a missing value.
According to the assumption of the primary hypothesis, H01a, no significant achievement of the load intensity of at least 40% (M = 104.53 ± 4.56) could be determined for both VR training programs. Regarding the VR-SET, there were no differences to the load limit of 40% in the “main part” (n = 22, M = 106.06 ± 19.22, paired t-test p = .688, 95% CI [-9.15, 6.16], d = -.087) and even significantly lower load values were achieved in the “main part” of the VR-ET (n = 21, M = 96.98 ± 15.16, paired t-test p = .017, 95% CI [1.53, 13.68], d = .570). Consequently, the H01a must be accepted. However, a comparison between the VR-SET and the VR-ET showed that significantly more participants reached the desired exercise intensity of 40-60% in the “main part” of the VR-SET than in the VR-ET (n = 21, Wilcoxon test p = 0.014, r = -.534). Thus, hypothesis H01b can be rejected.
In addition, various group comparisons were made to assess the load intensity during the different training sections. Within the group "training frequency" it was descriptively determined for both the VR-SET and the VR-ET that participants who train less frequently (0-1x per week) have higher HR values than participants who train more frequently (2x and more per week). Higher HR values were also determined within both training programs for the group of participants under 76 years of age and women, as well as for the group that had no previous VR experience. ANOVA with repeated measures found no statistically significant main effects between the groups for both training programs (Fig. 7). Hypothesis H01c must be accepted.
Fig. 7. Main Effects of ANOVA with repeated measures for groups in VR-SET and VR-ET.
Perceived exertion
The subjective perceived exertion was determined after each VR training session using Borg’s RPE scale. The RPE ranged from "easy" to "somewhat difficult" in the VR-SET (M =12.14 ± 1.81) and "easy" in the VR-ET (M = 10.82 ± 2.69).
When comparing the converted RPE values with the actual measured HR values during the “main part”, the values of RPE were higher than those of the objective measurement in both visits (Table 4). Contrary to hypothesis H02a, a significantly higher subjective perceived exertion, compared to the objectively determined exertion from the HR, was found during the “main part” of the VR-SET. With regard to the VR-ET, there were no differences between subjective perception and objective measurement. Accordingly, H02a can be assumed for the VR-ET. Likewise, H02b must be discarded due to the statistically significant difference between the two training sessions in the RPE (n = 22, paired t-test p = .028, CI 95% [1.54, 24.82], d = .502).
Table 4. Comparison of subjective perceived and objectively measured load.
|
Borg Scale
M (SD), bpm
|
HR
M (SD), bpm
|
p-value
|
95 % CI
|
d
|
VR-SET (n=22)
|
121.36 (18.07)
|
106.03 (19.22)
|
.010**
|
-26.62, -4.05
|
-.602
|
VR-ET (n=21)
|
110.48 (25.19)
|
96.98 (15.16)
|
.052
|
-27.15, .15
|
-.450
|
Paired t-test, M mean, SD standard deviation, HR heart rate, CI confidence interval, **p < 0.01
|
Blood pressure
The evaluation of BP was based on systolic and diastolic BP, as well as HR. Systolic baseline values were higher in the case of the VR-SET than the VR-ET before training (“pre”) (Fig. 8). Over the course of the three post-exercise measurements (“post”, “5 min post” and “10 min post”), systolic BP decreased for both the VR-SET and the VR-ET (Fig. 8.). In the VR-SET, the tests of within-subjects effects revealed a significant main effect of training on systolic BP (ANOVA with repeated measures, Greenhouse-Geisser correction, p < .001, partial η2 = .343, d = .722). When systolic BP was measured before and after VR-ET, similar results were obtained (ANOVA with repeated measures, Sphericity assumed, p < .001, partial η2 = .551, d = 1.108). The pairwise comparisons between the time points of each training session can be seen in Table 5. Accordingly, for systolic BP, H03a can be rejected in relation to the VR-SET and the VR-ET. Moreover, 5 min after training, the systolic BP was significantly lower in the case of the VR-ET than the VR-SET (paired t-test p = .020, CI 95% [-11.80, -2.51], d = .535). Regarding the time point “5 min post”, H03b has to be rejected.
Table 5. Pairwise comparisons of systolic BP in VR-SET and VR-ET.
(I)
|
(J)
|
MD (I-J)
|
SD
|
p-value
|
95 % CI
|
VR-SET (n = 22)
|
pre
|
post
|
.53
|
3.02
|
1.000
|
-8.25, 9.31
|
5min post
|
10.57
|
3.07
|
.015*
|
1.62, 19,51
|
10min post
|
9.55
|
2.85
|
.018*
|
1.27, 17,84
|
post
|
5min post
|
10.04
|
1.57
|
.000***
|
5.48, 14,60
|
10min post
|
9.02
|
1.89
|
.001**
|
3.52, 14,53
|
5min post
|
10min post
|
-1.02
|
1.61
|
1.000
|
-5.69, 3.66
|
VR-ET (n = 22)
|
pre
|
post
|
3.21
|
1.98
|
.718
|
-2.55, 8.98
|
5min post
|
9.75
|
2.35
|
.003**
|
2.89, 16.61
|
10min post
|
16.07
|
2.34
|
.000***
|
9.24, 22.89
|
post
|
5min post
|
6.54
|
1.52
|
.002**
|
2.12, 10.96
|
10min post
|
12.86
|
1.87
|
.000***
|
7.43, 18.29
|
5min post
|
10min post
|
6.32
|
1.74
|
.009**
|
1.26, 11.38
|
ANOVA, MD mean difference, SD standard deviation, CI confidence interval, *p < 0,05, **p < 0,01, ***p < 0,001
|
For diastolic BP, a decrease in BP values after the training session was noticeable in both training programs (Fig. 8.). Although significant, within-subject effects were determined for the VR-SET (ANOVA with repeated measures, Sphericity assumed, p < .046, partial η2 = .118, d = .366) and VR-ET (ANOVA with repeated measures, Greenhouse-Geisser correction, p = .002, partial η2 = .203, d = .505), a significant difference between the time point "post" and "10 min post" was only detected for the VR-ET in the pairwise comparison (MD = 4.19 ± 1.30, p = .025, CI 95% [.40, 7.98]). Based on the significant main effects, H03a can be rejected for both training sessions. Moreover, no significant difference could be found between the two training programs at all four time points. Therefore, H03b is retained for diastolic BP.
Whereas within the VR-ET the HR was, on average, almost constant between the time points (Fig. 8.) and thus no significant, within-subject effects could be recorded (ANOVA with repeated measures, Greenhouse-Geisser correction, p = .079, partial η2 = .101, d = .335), in the case of the VR-SET, internal significant main effects were found (ANOVA with repeated measures, Greenhouse-Geisser correction, p < .001, partial η2 = .314, d = .677). The pairwise comparison for the VR-SET showed differences between the time points "pre" and "post" (MD = -6.80 ± 1.65, p = .003, CI 95% [-11.62, -1.99]) as well as "post" and "10 min post" (MD = 3.93 ± 1,24, p = .028, CI 95% [.32, 7,54]). This leads to an acceptance of H03a in the case of the VR-ET and a rejection of the VR-SET. HR also differed between the VR-SET and the VR-ET. Accordingly, with a similar initial HR in both training sessions, a greater increase in HR was recorded for the VR-SET. This change resulted in significant differences between the training sessions at the time point "post" (paired t-test, p = .004, CI 95% [3.34, 10.66], d = .693), time point "5 min post" (paired t-test, p = .001, CI 95% [2.60, 8.69], d = .821) and time point "10 min post" (paired t-test, p = .002, CI 95% [2.15, 8.27], d = .752). Thus, H03b was rejected.
Fig. 8. Progression of BP values at the respective measurement times.
Perceived task load
In terms of the perceived task load during VR training (Fig. 9), mental demands were significantly higher in the VR-ET than in the VR-SET (n = 22, Wilcoxon test p < 0.001, r = -.800). In contrast, the VR-SET imposed significantly higher physical demands (n = 22, Wilcoxon test p = 0.050, r = -.419). No differences were found with respect to temporal demands (n = 22, Wilcoxon test p = 0.298, r = -.221). The effort expended to complete the tasks/exercises was significantly higher in the VR-SET than in the VR-ET (n = 22, Wilcoxon test p = 0.011, r = -.558).
Fig. 9. Boxplot of the adjusted task load values of the VR-SET and VR-ET.
In addition, the majority of participants found both the total duration (14 participants in the VR-SET and 13 participants in the VR-ET) and the break times (17 participants in the VR-SET and 21 participants in the VR-ET) of the exergame appropriate. Eight participants in the VR-SET and nine participants in the VR-ET, respectively, indicated that the total duration could be longer. In each training session, one person found that the breaks were too long. In the VR-SET, four people reported that the break times were too short.